Development of a novel web application for automatic photovoltaic system performance analysis and fault identification

This paper details the development of the pvpaR (PV Performance Analysis in R) web application built on open-source technologies for the automated and user-friendly evaluation of photovoltaic (PV) system performance and identification and classification of faults. pvpaR's ecosystem is based on the R statistical computing project, both for the back-end as well as for the front-end which uses the Shiny web application framework. Currently, the core of the application incorporates models for synthesizing time-series of irradiance, module temperature, PV system voltage, current, DC and AC power. These are used to validate the imported field measurements and create the comparison which the fault identification function is dependent upon. The web application can currently import measurements from flat files and databases. It has been released under the GNU Affero GPL license to encourage contributions, with the goal to become a useful tool for the PV community and PV system owners.